Uracil is a normal base in RNA but a miscoding lesion in DNA. Incorporation of uracil in the genome can occur through deamination of cytosine and also through occasional use of dUTP instead of TTP during DNA replication resulting in premutagenic U:G or U:A base pairs. Unrepaired uracil in DNA mispairs with adenine resulting in mutagenic phenotype that is potentially carcinogenic. To avoid the mutagenesis associated with unrepaired or insufficient repair of uracil, most organisms harbor uracil DNA glycosylase (UDG) in their cells. This enzyme is encoded by the ung gene in the nuclear genome but a splice variant is translocated to the mitochondria of higher organisms. Ung-/- mice have a modest increase in spontaneous mutation frequency (Nielsen et al., Mol. Cell, 2000, 5: 1059-1065). Quite recently, Nilsen and co-workers showed that ung deficient mice have increased incidence of B-cell lymphomas at old age (Nilsen et al., Oncogene; 2003, 22: 5381-5386). These data provide the first evidence of cancer development due to a deficiency in a DNA glycosylase in mice model in addition to a possible role of ung in the immune system. Furthermore, the data by Nilsen and co-workers also provide a link between carcinogenesis, DNA repair and the aging process. The UDG is a highly ubiquitous DNA repair enzyme that initiates the repair of uracil through the versatile DNA repair pathway known as base excision repair (BER). The major steps in BER include: scission of the bond between the inappropriate base and the sugar (glycosylic activity) by a DNA glycosylase leaving an apurinic/apyrimidinic (AP) site, phosphodiester bond cleavage by AP lyase activity of the same enzyme or by an AP endonuclease (AP lyase activity), addition of the correct nucleotide by DNA polymerases (polymerization) and ligation by DNA ligases. The polymerization step has been shown to diverge into two sub pathways: the short- and the long patch. In the short-patch BER sub pathway, only one nucleotide is incorporated after the glycosylic step, whereas, in the long-patch sub pathway, more than one nucleotide is incorporated. To date, the in vivo significance and regulation of the two BER sub pathways in the repair of uracil remains unclear. We have hypothesized that uracil repair in the mitochondria is accomplished via the short-patch mechanism and that the sub pathways of BER are largely determined by the nature of DNA glycosylases involved. Our preliminary results obtained using mitochondrial extracts of human lymphoblastoid origin suggest that uracil repair in the human system is accomplished exclusively via the short-patch BER sub pathway. In order to understand the mechanism of uracil BER in the mammalian mitochondria, we have engaged in a collaborative study with Dr. Samuel H. Wilson's laboratory of Structural Biology at NIEHS. In this study, we are using wild type and UDG knockout mouse fibroblasts, which were established by Dr. Wilson's group and oligonucleotides containing a single uracil at specific location. This oligonucleotide is designed so as to contain Hpa II and Hae III restriction digestion sites to help discern between short- and long patch BER sub pathways. Using this model system, we are assessing glycosylic activities of mitochondrial and nuclear isoforms of UDG. In addition, we are examining the mechanism of nucleotide incorporation (repair synthesis) during uracil BER. We are also studying the size (nucleotides) of the repair patch generated in the course of uracil repair. Furthermore, we are using this system to determine the nature of protein-protein interactions involved during the repair of uracil in mouse system. Since UDG is a pure DNA glycosylase without an associated AP lyase activity, it must perform the glycosylic bond scission and then hand over the resulting AP site to an AP endonuclease for further processing in order to complete the repair process. This notion would be consistent with the "passing the button" model proposed some years ago by Dr. Wilson. To ascertain if this is the case, we intend to perform repair synthesis reactions using cell-free extracts from mouse UDG-knockout and wild type cells in the presence of purified AP endonuclease. Proficient repair synthesis is expected in the presence of both UDG and AP endonuclease if the proposed model is true. However, the lack of UDG in the knockout cells may not support proficient repair synthesis even if AP endonuclease is present. This project may potentially allow us to determine and identify novel back-up DNA repair pathways for uracil in the mouse system.